RESISTANCE DEVICES 133 



swells and stresses the wire. The effect is most noticeable when 

 fine wires are used; that is, in high resistance coils. It follows 

 the seasonal variations of atmospheric humidity, and may be as 

 great as 0.04 per cent, in a 1,000-ohm coil. 



Immersion of the coil in oil which is freely exposed to the air, 

 though retarding the effect, is not an absolute preventive, for 

 the oil absorbs moisture and imparts it to the shellac. To over- 

 come this source of error, Rosa has developed the form of sealed 

 resistance standard 3 shown at B in Fig. 59. 



The coil itself is prepared as specified by the Reichsanstalt, 

 being insulated with silk and shellac and thoroughly dried by 

 baking; the bobbin is supported from the cover by the thermome- 

 ter tube. The case is nearly filled with a pure oil which has 

 been carefully freed from moisture. The cover is then screwed 

 into the protective case and the joint sealed with shellac. Poten- 

 tial terminals are used with coils of 1 ohm or less. Sealed stand- 

 ard coils show no seasonal variations in resistance. 



Fig. 59C shows a current-carrying standard. Such resistances 

 are immersed in oil, which is stirred and kept cool by a water 

 jacket through which there is a brisk circulation. It is essential 

 that the strips of resistance material be protected by a coating 

 of shellac. 



Resistance Coils for Alternating-current Work. Though the 

 bifilar winding usually employed in resistance coils reduces the 

 inductance to a minimum, it increases the possibility of capacity 

 effects, since at the terminals of the coil the two wires are separ- 

 ated by only the double thickness of insulation and have full 

 voltage between them; the P.D. between the wires decreases 

 with the increase of the distance from the terminals, becoming 

 zero at the end of the bight. 



It is clear that with alternating currents, especially at high 

 periodicities, the behavior of the coil will be modified by its 

 distributed capacity and inductance. The high resistivity of the 

 wires eliminates trouble from skin effect, which at 3,000 cycles 

 with a manganin wire 2 mm. in diameter is only about 1 part in 

 100,000. 



Assuming as an approximation that the formulae for two paral- 

 lel wires apply in this case, Curtis and Grover have deduced the 

 following relations 7 : 



